Statistical wormhole detection for mobile sensor networks

Abstract

A wormhole attack is one of the most challenging yet detrimental security issues in mobile wireless sensor networks (mWSNs). However, as most of the existing countermeasures are designed mainly for fixed WSNs using hardware devices or information of entire WSNs (topology or statistical), they cannot be effectively used in mWSNs. In this paper, we propose, Statistical Wormhole Apprehension using Neighbors (SWAN), a novel wormhole countermeasure for mWSNs. As SWAN utilizes the localized statistical neighborhood information collected by mobile nodes, it apprehends wormholes not only without requiring any special hardware device but also without causing significant communication and coordination overhead. We performed extensive studies on false positive and detection rates via both analysis and simulations. Our simulation results show that SWAN can detect wormhole attacks with high probabilities and very low false positive rates.